Men at work: Mobile EEG as a new tool for neuroergonomics

Our recent study “Age-sensitive effects of enduring work with alternating cognitive and physical load. A study applying mobile EEG in a real life working scenario” (Wascher et al., 2016) demonstrates how mobile EEG can be utilized not only for user state examination but also for the evaluation of cognitive processing in real life working situations.

The proportion of older employees is increasing continuously. Thus, the adjustment of workplaces to the needs of the older population becomes increasingly important. Ergonomic assessments of physical aspects of work primarily use observation and surveys. Cognition, however, and human factors affecting cognition, can be addressed only superficially in real life situations because they are not reliably observable this way. Neurophysiological measures applied while regular work is performed may help to close this gap.

Mental states like fatigue, motivation loss, or stress and their impact on cognitive processing are well-known risk factors at the workplace. There is a long history of laboratory research that intends to provide reliable measures for these states without interfering with the primary task. There are numerous EEG studies (e.g., investigating drivers’ fatigue; Lal & Craig, 2001) that reported an increase of power in lower frequency bands (Alpha, Theta) with increasing mental fatigue (Aeschbach et al., 1997; Kiroy et al., 1996). Based on these data, oscillatory EEG activity has even been proposed as a potential countermeasure for mental fatigue at the workplace (Lal & Craig, 2001).

However, these types of studies have two shortcomings: Firstly, laboratory settings are in most cases much less complex than real life scenarios. Secondly, only global states of mental states can be addressed using these methods. Specific aspects of human information are lost, as no distinct stimuli (required for investigating event-related EEG activity) can be set without interfering with the task at hand.

Assuming that eye blinks denote kind of segmenting events for the visual input stream (Doughty, 2001; Wascher et al., 2015), here we used the eye-blink-related activity of the EEG to analyse cognitive processing in a real life working situation. Simulating a workplace at a wholesale dealer, older and younger adults performed a regular work shift, wearing mobile EEG.

Data Analysis

Eye blinks were used as temporal markers for EEG analyses. Data segments from -1000 to 2000 ms around the blink maximum in the vEOG were extracted (baseline -450 to -250 ms). An independent component analysis (ICA) was applied for the removal of artifacts.

For analyses of frequency spectra, fast-fourier transformations (FFTs) were applied. Spectral power was calculated based on individual Alpha frequency (IAF; Klimesch, 1999), with mean Alpha power between IAF-2 to IAF+2Hz, and mean Theta power from IAF-5 to IAF-3. Event-related desynchronisation and synchronisation (ERD/ERS) analyses were performed according to Pfurtscheller & Aranibar (1979) for both frequency bands. Mean power between 0 and 300ms after the re-opening of the eyes was calculated for analyses.

Signal analyses were performed on MATLAB® using EEGLAB (Delorme & Makeig, 2004), statistical analyses were conducted using GNU R (R Development Core Team, 2012), and plots were drawn using VEUSZ (Sanders, 2013).

Self Assessment

Fatigue increased with time-on-task (ToT), F(2,48) = 9.70, p = .001. A clear modulation was found with the task performed, F(3,72) = 56.04, p < .001: In particular, after the Simon task mental fatigue was high. The ToT effect was more pronounced in older participants, F(2,48) = 3.33, p = .061. Additionally, some evidence for an interaction of age by task was found, F(3,72) = 2.44, p = .108, indicating higher fatigue in the Simon task in younger, than older, adults.

Behavioral Data (Simon Task)

Response times (RTs): Older adults responded marginally slower than younger participants, F(1,23) = 3.99, p = .058. No overall effect of ToT was found, however, within blocks, RTs were faster after the physical task, F(1,23) = 7.67, p = .007.

Errors: No age effect was found, but error rates slightly increased with ToT, F(2,46) = 5.16, p = .013. This effect was more pronounced in younger participants, F(2,46) = 2.79, p = .079, indicating that they committed more errors later in the experiment

EEG Data

Alpha power (see figure 2) increased with ToT, F(1,24) = 17.76, p < .001, and varied with task, F(2,48) = 10.96, p < .001. It was higher in the monotonous Simon task than in the self-paced D2 task, F(1,24) = 15.40, p = .002, and higher when participants performed the physical compared to the self-paced D2 task, F(1,24) = 18.18, p < .001. ToT modulated the effect of task, F(2,48) = 6.97, p = .004, with larger ToT effects in the cognitive tasks. The increase of Alpha power in the Simon task was restricted to younger adults, F(2,48) =  5.81, p = .012.

Figure 2: Mean Alpha power (with standard error of mean), separated for tasks and EEG channels. Besides a general effect of ToT, younger adults showed a massive increase in Alpha power in the monotonous Simon task.

Theta power was reduced in older adults, F(1,24) = 5.29, p = .030, and varied with the task performed, F(2,48) = 26.64, p < .001. Theta power was markedly increased in the physical task compared to the self-paced cognitive D2 task, F(1,24) = 37.66, p < .002. No effect of ToT was observed. Significant age effects were only observed in the cognitive tasks, D2: F(1,24) =  8.96; Simon: F(1,24) = 10.48, both p < .05.

Alpha ERS (see figure 3) was enhanced for older adults, F(1,24) = 9.81, p = .005, and varied with the task performed, F(2,48) = 6.97, p = .002. ERS systematically increased with ToT in younger adults, F(1,12) = 11.25, p = .012, but not in older ones, F(1,12) = 1.49, p = .492.

Figure 3: Time course of Alpha synchronization for the three tasks and two EEG channels, superposed for the 3 Blocks in the experiment (BL1, BL2, BL3). Data are time locked to the maximum of vEOG activation during the blink. Most prominently, Alpha synchronization was more pronounced in older adults.

Theta ERS was enhanced in older adults, F(1,24) = 7.52, p = .011, and varied with the task performed, F(2,48) =  3.00, p = .059. Theta ERS was slightly higher in the Simon task compared to the self-paced D2 task, F(1,24) = 4.31, p = .049, but did not differ between the two self-paced tasks, F(1,24) =  0.21, p = .652.

In sum, the pattern of EEG data is well comparable to previous laboratory settings. Alpha power increased with ToT (see Wascher et al., 2014), but showed a marked reduction in older adults, however, only in the monotonous cognitive task. Theta power was reduced in older adults before most in cognitive tasks. Strongly enhanced synchronization of both frequency bands was observed in older adults.

The present study simulated a realistic working shift in the post room of a wholesales house. Participants moved parcels and performed cognitive tasks on the computer that were either repetitive (and rather monotonous) or self-paced. During the entire shift, the participants’ EEG was recorded by mobile EEG equipment that did not restrict free movement and thus allowed natural behavior at any moment.

Subjectively experienced fatigue remained rather stable in younger adults, but increased in older adults. For both groups, however, fatigue was highly related to the task performed. After the monotonous computer task, fatigue ratings were increased compared to the physical task, a finding that nicely stresses the role of monotony for the experience of mental fatigue.

The enhanced impact of monotony upon younger participants was also mirrored in increasing error rates (but not RTs) with ToT. Thus, at the end of the working shift, their accuracy in the Simon task was even lower than that of the older participants. Also, younger participants showed markedly increased alpha activity in this particular task. Assuming high alpha power to reflect an idle state of the attentional system (Hanslmayr, 2012), younger adults might have switched to a state of attentional withdrawal (Wascher et al., 2014).

Older adults appeared to deal better with monotony, and factors of fatigue were more widespread across tasks. In general, age is related to larger cortical phase synchronization (Müller, et al. 2009). In particular early EEG synchronizations might indicate that older adults were more driven by external signals (see Klimesch et al., 2002). These data are in accordance with laboratory studies, indicating amplified early EEG responses in older adults. This stronger impact of stimulation might be due to reduced executive cognitive control observed with increasing age (Gazzaley et al., 2008). As a neurophysiological correlate, reduced frontal theta activity has been discussed (Cummins & Finnigan, 2007), as also found in the present study. However, this effect was restricted to the cognitive tasks, suggesting that the age-related decrease in theta power is not a global fact, but rather task-dependent. Most interestingly, ERD/ERS showed convergence between age groups with ToT. In younger participants monotony led to an impairment of executive control functions and to more stimulus-driven behavior, and thus to a transient state that resembled the aging brain.

In sum, applying neural measures to a real life work situation provided substantial information about mental fatigue in younger and older adults. The diversity of tasks additionally provided important insight into the meaning and usefulness of neurophysiological measures for neuro-ergonomics. Most importantly, blink-related activity in the EEG was systematically related to the task performed and to other experimental factors. Event-related EEG analyses without any external stimulation appear especially suited in working situations, as nothing has to be changed to the natural environment. Thus, mobile EEG provides substantial information about cognitive processing at the workplace and its alteration due to fatigue or age-related aspects.